Integrating the voluntary sector in personalised care: mixed methods study of the outcomes from wellbeing co-ordination for adults with complex needs

Purpose – This integrated care study seeks to highlight how voluntary sector “wellbeing co-ordinators” colocated in a horizontally and vertically integrated, multidisciplinary community hub within one locality of an Integrated Care Organisation contribute to complex, person-centred, co-ordinated care. Design/methodology/approach – This is a naturalistic, mixed method and mixed data study. It is complementing a before-and-after study with a sub-group analysis of people receiving input from the wider hub (includingWellbeing Co-ordination and Enhanced Intermediate Care), qualitative case studies, interviews, and observations co-produced with embedded researchers-in-residence. Findings –The cross-case analysis uses trajectories and outcome patterns across six client groups to illustrate the bio-psycho-social complexity of each group across the life course, corresponding with the range of inputs offered by the hub. Study of outcome from wellbeing co-ordination


Introduction
The definition of health adopted by the World Health Organisation (WHO) in 1946 as "a state of complete physical, mental and social wellbeing and not merely the absence of disease or infirmity" is as timeless as it is timely. It serves as a continuing reminder to practitioners, policymakers and academics to work together to implement integrated (or person-centred, coordinated (Lloyd et al., 2017) services that are informed by an underlying bio-psycho-social model and approach. While the concept of integrated care has been developed over time, there has been less focus on the approach required to underpin an ethos of the whole person, as opposed to disease-specific care, e.g. . The core challenge is to transform care systems from reactive, deficit-based models to anticipatory, holistic, strengths-based approaches that are focussing on capacity and functioning (World Health Organisation, 2017).
As individuals live longer, but within a world where different forms of social adversity continue, the proportion of the population with complex bio-psycho-social needs is increasing. This is particularly so for those older individuals living with multi-morbidity, frailty and dementia (Cassell et al., 2018), those with needs related to homelessness, substance misuse and mental health problems (Balabanovic and Hayton, 2019), and those who have palliative and end of life care needs.
These complex groups require a range of services from prevention, rehabilitation, palliative care and bereavement support across community, primary and secondary care, physical and mental health care and across health, social and informal care.
Lessons from social prescribing Under the banner of "social prescribing", shifting the care of individuals from a clinical to a more social and person-based model (in part by linking them to interventions outside of healthcare) has become the "Zeitgeist" in UK health policy (NHS England, 2018). Indeed, social prescribing is a cornerstone of the new NHS Universal Personalised Care model (Sanderson et al., 2019). There is a financial commitment for every Primary Care Network (PCN) to have a social prescribing link-worker, funding for new workers becoming available from 2019 with the goal of having 1,000 new workers in place by 2020/2021. Link workers will become an integral part of the multi-disciplinary teams aligned with PCNs, which also include pharmacists, physician associates, physiotherapists and paramedics. They also provide a key understanding of and links to the voluntary, community and social enterprise sector (VCSE), the main provider of non-clinical practical and emotional support services.
Although social prescribing could not exist without the infrastructure of the VCSE for link workers to refer to, the contribution of the VCSE has been poorly researched and underreported (Department of Health and Social Care/Public Health England/NHS England, 2016; Lovell et al., 2017;Husk et al., 2019;Smith et al., 2019). This is despite the fact that the UK Government itself (Center for Workforce Intelligence, 2015) acknowledges that the everincreasing burden of care will continue to be held by the communities by themselves, with unpaid and voluntary care accounting for 61% of the total time of health, social care and public health activity.
Insofar as the introduction of social prescribing formally recognises what the VCSE has been doing for years and is part of a national programme which is keen to demonstrate and evidence its impact, the roll out of social prescribing opportunities provide an opportunity to explore the role of the VCSE in supporting integrated, personalised services. To this end, this paper reports a sub-group cross-case analysis of a larger evaluation of a complex integrated care programme in Torbay and South Devon (TSD) and, more specifically, in one of five sublocalities, Coastal. This locality was the first in TSD to co-locate voluntary sector link workers with intermediate, community-based services in order to provide a holistic service, a similar approach to that explored by Dickinson and Neal (2011). Drawing on mixed methods, we analyse outcomes from wellbeing co-ordination for a cohort of adults with needs that range across the care pathway, from prevention to bereavement support.

Setting
Torbay and South Devon NHS Foundation Trust provides acute and community services organised around five localities, ranging in size from 35,000 to 72,000 registered residents. It is an Integrated Care Organisation (ICO), one of the first in England, and formed out of a merger between the acute NHS Trust and Torbay Care Trust in October 2015. The area has a strong history of horizontal service integration, following the creation of the Torbay Care Trust in 2005 (Thistlethwaite, 2011;Wilding, 2010).
From 2016, several changes were made to the organisation and delivery of care in the ICO to promote responsive and dependable out-of-hospital services that provide high-quality assessment and support to people in their own homes. Five of the nine community hospitals in four localities (the largest, Torquay, houses an acute hospital) were closed from March 2017, resulting in a reduction in the number of community hospital beds. In their place, intermediate care teams (comprising nurses and therapists) were "enhanced" to include General Practitioners (GPs) and more recently pharmacists, and voluntary sector "Well-Being Coordinators" (WBCs); with "Health and Well-Being Centres" established (some in former hospital sites).

The coastal locality
The "Coastal" locality of interest for this study covers the Coastal market towns of Teignmouth and Dawlish and their surrounding areas. It has a high proportion of older people compared to England (22% of the 36,251 Coastal locality residents are over 70 years). Despite developing the same Care Model, performance (as measured by higher Intermediate Care referral rates, lower lengths of episodes for intermediate care, higher rates of homebased care, and lower GP referrals to Accident and Emergency (A&E)) appears to be better in Coastal compared with the other localities, with differences that reach statistical significance Elston et al., 2019a, b). The fact that Coastal was an earlier adopter of the new Care Model than other localities may, in part, explain its better performance. Using the Study of outcome from wellbeing co-ordination "Context and Capabilities for Integrating Care (CCIC) Framework" (Evans JM et al., 2016), physical features (structural and geographic) were found to be very pertinent to the relative success of the Coastal Locality, as were empowered clinical leadership, with readiness for change being expressed through processes and cultures that were risk-enabling, strengthsbased, person-and outcome-focussed .
Multi-disciplinary team working An Enhanced Intermediate Care (EIC) team comprising local GPs (five sessions per week), a pharmacist (four sessions per week) and two part-time Wellbeing Co-ordinators (voluntary sector representatives) meet as a Multi-Disciplinary Team (MDT) with community matrons, nurses, occupational therapists and physiotherapists, social workers, mental health liaison staff, and health and social care co-ordinators. The MDT meets seven days a week to discuss and plan the care for approximately 30 high-risk or newly discharged patients each day. The EIC team are co-located within the Teignmouth Health and Wellbeing hub (a former community hospital), with strong links to the wider community team (based in the same building), the community hospital based at Dawlish, and the voluntary sector partner organisation establishing their offices and a walk-in reception area.

Voluntary sector link workers
Wellbeing Co-ordinators (non NHS staff) were commissioned by the ICO, as part of a wider social prescribing initiative. Local GPs were actively encouraged to refer deteriorating older patients to EIC and WBC. The two Wellbeing Co-ordinators use goal-setting tools to produce a strengths-based, person-centred plan. They work with individuals (and their families/ caregivers) for up to 12 weeks to enable them to take action to achieve their goals. Support includes resilience focussed one-to-one coaching and practical support and advocacy to navigate and access other local statutory and voluntary services. To this end, they received training in goal-setting, use of tools and outcome measures and how to engage with clients in a strengths-based way, co-produce a plan, and manage risk.

Methods
A before-and-after study was published on the wider WBC programme measuring health and social well-being, activation and frailty at 12 weeks and social care, community and secondary care service use and cost at 12 months prior and after the intervention (Elston et al., 2019a, b). Quantitative findings showed that most of the 86 participants achieved their goals (85%) and that on average health and well-being, patient activation and frailty showed a statistically significant improvement in mean score. Mean activity increased for all services. Forty-four per cent of participants saw a decrease in service use or no change. Thirteen highcost users (>£5000 change in costs) accounted for 59% of the overall cost increase. This was largely due to significant, rapid escalation in morbidity and frailty, and inspired the subgroup analysis for the purpose of this paper below. Due to the small sample, incomplete data, and lack of control, group cost patterns were used in this paper to highlight the complexity of the subgroups and help in building a chain of evidence.
The evaluation above is supplemented here with a sub-group cross-case analysis, 7 qualitative patient interviews, 3 in-depth, un-structured voluntary sector practitioner interviews, 15 observations of multidisciplinary team meetings (9 daily hub meetings, 4 weekly voluntary sector triage/referral meetings, 2 monthly voluntary sector core team meetings), and 3 observations of service user/caregiver interactions over a three-year period. Observation data was largely opportunistic and conducted by participant observers (i.e. researchers-in-residence).

JICA 28,4
Secondary qualitative client data-care records and case notes For an in-depth cross-case comparison within sub-groups of clients, additional secondary data was drawn from clients' care records, as well as practitioner case notes (all 49, including N 5 20 who also received Enhanced Intermediate Care support), and in-depth case notes compiled by wellbeing co-ordinators (N 5 16).

Qualitative data analysis
Qualitative data were recorded with audio and field notes, transcripts of which were uploaded into an NVIVO file, where they were merged for sub-group analysis with the quantitative survey responses and other documentary materials (case notes, case studies and care record extracts).
For the purpose of cross-case comparison, survey data and case-based qualitative data were coded up using basic thematic analysis (Braun and Clarke, 2006), identifying the key pathways and core functions, as well as identifying the underlying bio-psycho-social needs, as well as inputs and outcome patterns for each case.
This in-depth study is informed by a responsive evaluation approach to supplement the summative evaluation in which we look at the system, team and individual clients (Stake, 2004), explore the particular and the personal and combine questions around "what works" with "how things work" and "what is going on".
Secondary quantitative outcome measures Data on age, sex, referral route was recorded on referral forms and health outcomes collected at either the first or second meeting with the co-ordinator and after 12 weeks of support (or the point of exit). As part of the NHS commissioning contract voluntary sector wellbeing service providers were obliged to collect the following validated health outcome measures: Outcomes Star Motivational Chart Tool, Age UK (seven domains, each scoring 1 to 10), Patient Activation Measure (PAM) (13 items, scoring 0 to 100, or levels 1 to 4) and Warwick Edinburgh Mental Health and Wellbeing Scale (WEMWBS) (Tennant et al., 2007) (14 Likert questions, each scoring 1 to 5) and Rockwood Clinical Frailty scale (RCFS) (7 domains, scoring 1 to 10) (Rockwood et al., 2005). These outcome measures were selected by commissioners to reflect a wide and robust range, reflecting personalised and comparable outcomes adopted and endorsed nationally (like PAM, WEMWBS). These and counting goals and their attainment were intended to reflect the core tenet of the care model, i.e. what matters to the patients and were used to inform care planning. The frailty scale is widely used clinically across NHS settings locally and reflects an additional more bio-medical marker to include. As this is research based on secondary evaluation data, there were no resources in the system to consider comparisons with a more centralised frailty marker like the Electronic Frailty Index or Devon Predictive Model.

Ethical considerations
Individual consent was obtained from Wellbeing Co-ordinators collecting outcome data from questionnaires, while ethical approval to use evaluation data as a secondary data source for research purposes was granted through "Proportionate Review" by the NHS Health Research Authority (Research Ethics Committee reference: 17/LO/1745; Protocol number: PSMD-208147-SA-FG-034; Integrated Research Application System project ID: 208147).
More broadly, the study is based on an action-based participatory approach, findings being co-produced in partnership with stakeholders and members of the public by embedded researchers-in-residence (Vindrola Padros et al., 2018). This took the form of co-re-design and implementation alongside the evaluation, as reported elsewhere .

The sample
In-depth analysis is conducted in this paper of n 5 49 people who benefited from the WBC services in the first six months of operation of the Wellbeing Co-ordination service in Coastal Locality. Of the 49 cases, 20 people also received one or more episodes of Enhanced Intermediate Care input in the 12 months before or after the WBC intervention. The sample included 33 female and 16 male clients. The age distribution was skewed towards older people, with a mean of 77.4 (median 80.7; range 52-99). The majority (N 5 37) were living alone, another six people were caregivers and another six people were being cared for (no overlap of these). Nine people experienced bereavement, and nine people deceased some time during or after the follow-up period. The groups overlap just in one case, where both the caregiver and the later deceased partner were supported simultaneously.

Identification of sub-groups
The main areas of focussed work by the Wellbeing Co-ordinators were identified from the 49 case records and 16 case studies compiled by the voluntary sector team. These areas of focussed work were coded up and ranked within a simple bio-psycho-social framework according to the main presentation. Based on this coding, as well as age, activity/cost patterns, and coding of trajectories, six sub-groups were identified (see Box 1).

Box 1. WBC subgroups
Maintaining mental health (n 5 5): The main presentation was identified as relating to significant mental health issues. Three of the cases had a history of suicidal ideation, two of whom were living with chronic pain; one homeless person was waiting for referrals to mental health services. It is important to stress that mental health was a prevalent issue across all other sub-groups. Caregiver/bereavement support (n 5 9): People in this sub-group came to the service already bereaved (n 5 4); were caregivers who were later bereaved (N 5 4); or was a caregiver/ volunteer in transition after significant health issues (n 5 1). Needing social support (n 5 11): Main presentation was identified as needing social support due to social isolation (n 5 6), housing/benefits issues (n 5 3) and family issues (n 5 2). The group also includes (n 5 2) in receipt of EIC. It is important to stress that loneliness and isolation were a prevalent issue across all other sub-groups. Mobilising (n 5 7): A mixed group of people that all had mobility issues due to, e.g. Parkinson's disease, anaemia, stroke, falls or falls risk. All were in receipt of EIC, some with re-referrals. Declining (n 5 10): People who were described as rapidly declining and complex, all of whom (except one who was referred from social care) had had previous EIC input. They had high service needs, 8 out of the 10 people incurring cost increases of over £10k from 12 months before to 12 months after the WBC intervention (range from £11,340 to £28,600). Ending (n 5 9): All of this sub-group have since deceased, with Wellbeing Co-ordinators serving in conjunction with Enhanced Intermediate Care (5 of 9, with one case being admitted after Enhanced Intermediate Care, 1 re-referral, and two cases with Enhanced Intermediate Care input 3 consecutive times over the follow-up period).

JICA 28,4
Categorisation of inputs Based on the analysis of records, case notes and case studies, Wellbeing Co-ordinators offered people in the sample 186 inputs in conjunction with related voluntary sector services (see Box 2).
Notably, all sub-groups were drawing from all these offers, ranging from 3 to 5 inputs per person. The mental health group drew most inputs (5pp), followed by caregivers (4.6pp), ending (3.88pp), mobilising (3.85pp), declining (3.6pp), and social mainly (3pp) groups. Inputs tended to focus predominantly on mental health support in the mental health and caregiver/ bereavement groups. Inputs were more practical in nature and included more care coordination for the other groups.
Grouping the inputs by priority areas of bio-psycho-social presentations, "social" inputs (n 5 47) were the largest area of focus, followed by "biomedical" (n 5 39), and "psychological" (n 5 36) inputs. The ranking of the prevalence of bio-psycho-social issues also shows increasing levels of complexity as the focus shifts from predominantly psycho-social (Mental Health, Caregiver/bereavement, social supporting groups) to more biomedical issues (Mobilising, Declining, Ending groups). For biomedical issues, the voluntary sector works more closely with the Enhanced Intermediate Care team, whereas psychological matters tend to involve more tertiary care. Most people had at least two, and around half (n 5 25) presented with all three categories; only one person in the social supporting group presented with a single social issue to work on.

Outcomes
The person-reported outcomes before and after the intervention are summarised in Table 1 and indicate that, on average, most people seemed to benefit meaningfully or at least remained stable following the intervention. Exceptions to this overall impact were two cases that were palliative on referral and where scores actually decreased on various outcomes, including mental health, outcomes star, frailty and activation. In three other cases, mental health declined up to 4 points due to ongoing chronic pain/depression, subsequent bereavement and rare disease (ultimately End-of-Life). Frailty scores also increased one point each in three people with relatively high entry levels (moving from 5 mild-6 moderate and an End-of-Life case moving from 8 very severe-9 terminally ill).
Care needs to be taken in interpreting the results. The sample size of the sub-groups is small. Activity data on primary care and mental health services activity are missing, and it is not possible to establish counterfactuals. Nevertheless, the data do indicate broad outcome patterns that suggest that the categorisation of sub-groupsbased on Wellbeing coordinator's own assessmentsdoes provide a plausible indication of the types of people that might most benefit from person-based social prescribing. Before and after intervention percentage increases in the Warwick-Edinburgh Mental Wellbeing Scale, the Patient Activation Measure and the Outcomes Star Motivational Chart Tool Age UK are relatively higher in the maintaining mental health group and, to a lesser extent, the caregiver/ bereavement support group. Other groups showed less evidence of benefit, although all  showed improvement in their outcome star measures, which may suggest the impact of WBC on self-efficacy. This is supported by the fact that all groups showed an improvement in goalsetting. Similar patterns can be seen with respect to activity and cost patterns. For example, the mental health group was the youngest and poorest in terms of baseline wellbeing scores and incurring high before-intervention costs. Post-intervention, this group not only experienced the greatest percentage increases in wellbeing scores (excepting the Rockford Clinical Frailty Scale). They also accounted for the greatest cost reductions in the year after. An opposite trend occurs with the declining and ending groups, who experience a decline in frailty and a significant increase in costs.

Capturing process factors
The results suggest that the WBC programme in the Coastal Locality has had an overall positive impact on individual wellbeing scores; and a significant impact on some sub-groups.
A key question that arises is whether voluntary input had a role to play in shaping these outcomes. Qualitative evidence suggests that this is the case. One advantage of voluntary sector staff, as referred to by the team, is their "acceptability" (Vol Sector team meeting, 24.01.18) over those from the statutory sector. In contrast to a possibly intimidating clinical environment ("it's in my space" Client interview 15.02.18). The voluntary sector team stresses that "we have time" and that they are non-threatening ("not wearing uniform") and engaging: "we go to their home, we offer face-to-face" (Voluntary Sector referral meeting 28.02.18).
Across qualitative interviews and observations, active ingredients in building the keyworking relationship include building trust, normalisation of suffering, feeding back progress and setting goals, mirroring emotions, involving the family and wider networks, connecting based on a common history of working together, nudging, setting long-term goals, and managing expectations around dependency.
Findings also suggest that achieving continuity of care provision and key working relationship requires time, resources and commitment, as well as a strengths-based approach ("we are not fixers but listeners", Wellbeing Co-ordinators double interview 24.01.18). Most people across sub-groups seem to be continuously monitored and in contact with voluntary services through various connection points, like volunteer cleaners, befrienders and drivers.

Discussion
This study delivers evidence that a co-located and integrated voluntary sector partner can play an important part in the mature, transitional care model implemented across one locality within a wider integrated care system. It demonstrates that complex care management from voluntary sector wellbeing co-ordinators makes a difference within an integrated care subsystem in terms of improving wellbeing outcomes and, for some groups, reducing health and social care costs.
The results also point to a generic need for holistic bio-psycho-social care and support approach that understands the complex needs of individual patients, aiding the bridging of key interfaces between primary, acute, mental health, social and tertiary care. The Coastal Locality suggests that WBCs can successfully fulfil this role and ensure that inputs are appropriately tailored to person-based needs.
It appears that goal-setting allied to personal outcomes have similar benefit across the continuum from prevention, rehab, palliation and bereavement support. This suggests potential for benefit from wellbeing co-ordinators extends to the palliative care and bereavement phase of chronic care. We think that it is interesting that Outcome Star scores improved across all sub-groups, while improvements in other measures (e.g. in the Warwick-Study of outcome from wellbeing co-ordination Edinburgh Mental Wellbeing Scale and the Patient Activation Measure) were more differentiated. It is difficult to comment on the relative validity of these measures, suffice to say that, in our experience, Outcome Star appeals to patients themselves, who recognise the multi-factorial influences on their well-being and who particularly enjoy the facility to visually map improvement in different dimensions. Just as social prescribing is designed to encourage a holistic approach based on the person's priorities, we may need to reconsider the way in which we conceptualise and measure wellbeing outcomes in social prescribing. The variable improvement in functional outcomes and activation across the different subgroups supports the need for earlier intervention and better identification of rehab and intermediate care input, and more nuanced outcome measures for people with advanced frailty and dementia at a palliative and end-of-life phase.
The paper also provides an exemplar around which difficult decisions may be made about implementing the universal personalised care package. It is important to note that, while both the declining and ending groups in this study experienced positive improvements in their Outcome Star scores, improvements with respect to the Warwick-Edinburgh Mental Wellbeing Scale and the Patient Activation Measure were less impressive and, in the case of the ending group, people experienced a decline in frailty. The average health and social care costs of these groups also increased significantly before and after an intervention. Given their poor clinical health status, this is not surprising. However, the emerging evidence of social prescribing stresses that in addition to benefits to individual patients and their carers, social prescribing is expected to lead to a decrease in health and social care costs (Munford et al., 2020). We consider this unlikely for the very frail and those ending their lives. Yet, if social prescribing can improve their sense of well-being and self-efficacy, should interventions be denied to these groups on the basis of lower costeffectiveness?

Strengths and limitations
This case study presents some evidence that a whole system intervention that mainstreams the role of the voluntary sector is having a positive impact on individual outcomes and may be having a positive impact on costs to the statutory sector. While we cannot definitely say that deploying voluntary sector workers as link workers was pivotal to improved outcomes, the qualitative elements of the research do suggest that VCSE link workers offer particular benefits.
It is also important to acknowledge that this sample was, overall, a costly cohort. We have not established a counterfactual to compare outcomes with usual care. Nor could we access activity and cost data on primary care and mental health services. The small sample size, particularly for sub-groups, is also a limitation. Of relevance to this special issue, we could also not establish the place of death for those people that have since deceased and could, therefore, not draw conclusions regarding cost implications (i.e. home, care home, hospital, or hospice). The data collection was opportunistic, conducted by participant observers (i.e. researchers-in-residence), and co-produced with NHS and voluntary sector representatives, possibly introducing a degree of bias.
On the other hand, the results of this study are plausible. This case study does not entail a control group. Thus, attribution is an issue. Further theorising around this would be beneficial. Nevertheless, one would expect the benefits of social as opposed to clinical interventions to be greater among groups suffering from psycho-social as opposed to biomedical problems. It may be important to bear this in mind when considering whom to refer to social prescribing problems, particularly those requiring end-of-life care.
Dis-/convergence with existing evidence. Rigorous studies on elements of integrating the voluntary sector in statutory services are mainly based within primary care settings JICA 28,4 (Tierney et al., 2020), whereas this was co-located with a community hub and mainly getting referrals from other routes (but including primary care). One trial in 2000 showed similar improvements in the intervention cohort with regards to psycho-social outcomes, while also showing an overall increase in cost (Grant et al., 2000). A scoping review in 2015 confirms this case study's findings around effects on people presenting with mental health and social isolation (Mossabir et al., 2015). More recently, larger mixed-method before and after studies of social prescribing co-ordinators in North of England (Woodall et al., 2018) and London (Carnes et al., 2017) are emerging and remain largely inconclusive. The further plan is to use current findings for a larger follow-up sample, and also include matched controls to fill this clear gap in the effectiveness evidence base of voluntary-sector based interventions (including helping define the active elements and underlying mechanisms of the plethora of varied programmes assumed under "social prescribing") (Tierney et al., 2020).
Reservations aside, we propose that the Wellbeing Co-ordination programme in the Coastal Locality of TSD demonstrates that integration between health services, social services and other care providers like the voluntary sector (horizontal integration) is possible and impactful; and that the voluntary sector can play a pivotal and positive role in integration. Moreover, understanding the cohort, in terms of their bio-psycho-social need, is critical to both adapting appropriate inputs and understanding variations in outcomes.

Acknowledgments
VIH team: Wellbeing Co-ordinators (Jeannette Osmond, Jill Breyley) for liaising/helping with consent for in-depth interviews/WBC-client þ referral observations/focus groups þ writing and co-presenting case studies þ in-depth interview/many informal debriefs; Administrator Miia Thomas for data retrieval on charity log/observation of VIH referral meetings, Volunteer Jackie for interview, Marj Dawson-home help manager observation of VIH referral meeting, Emma Flint -Volunteer and Befriending Co-ordinator for interview/referral meeting observation; Bob Alford Totnes Caring for implementation; Julia Street from Alice Cross for meetings.
Wider Hub and community: Louise Dawson Coastal Therapy Manager and Intermediate Care Lead for chairing MDTs/data retrieval on ICON system and case study liaising/consent; Health and social care co-ordinators for care record retrieval on ICON; Keri Dove Community Nurse Manager/Advanced Nurse Practitioner for dissemination, Werner Coetzee Devon County Council Team Manager Coastal Community Health and Social Care Team for dissemination, Victoria Coulson Lead Locality Medicines Optimisation Pharmacist for dissemination; Tammy Jones Senior Mental Health Practitioner for dissemination; Long-term conditions community matron Marcia Doherty co-ordinating three co-design workshops and dissemination; GP and Locality Clinical Director Mat Fox for championing and dissemination; GP Karen Bates co-ordinating three co-design workshops; Barrie Behenna chairing CCG engagement group; Jo Curtis CCG engagement lead for public engagement/ dissemination, Rebecca Harty CCG care model lead for engagement/dissemination, Ray Chalmers CCG Communication for engagement/dissemination; Jenny Turner CCG integrated care lead for engagement/dissemination; John Barry Stevens patient representative and PPG member.
System wide: Katie Chaplin Teign CVS for facilitating WBC learning events; Neil Elliott Information Manager for cost/activity data collection; Susan Martin QI Lead and research manager; Dawn Butler for executive sponsorship; Jo Smith PenARC research manager; Stephen Honeywill Associate Director for information governance; Catherine Williams AD for information governance; Helen Harman Age UK Torbay information governance; Sue McDermott Torbay Community Development Trust for information governance.

Study of outcome from wellbeing co-ordination
Funding: This work was funded/by the Torbay Medical Research Fund under Grant No 120, and the Torbay and South Devon NHS Foundation Trust (TSDFT). The research was supported by the National Institute for Health Research (NIHR) Applied Research Collaboration (ARC) South West Peninsula (PenARC). The views expressed are those of the author(s) and not necessarily those of the NHS, the NIHR or the Department of Health and Social Care. Contributor Statement: FG-qual data collection/analysis/writing/dissemination; JE-quant. data collection/analysis/writing/dissemination, SA-supervision and writing, CM qual. data collection/analysis/writing/dissemination; SW-project management and quant. data collection/analysis/writing/dissemination; RB-supervision and writing. Conflicting interest: The co-authors declare no conflicting interests, apart from those that might be assumed from the research being matched-funded by the NHS host and commissioning organisation of the voluntary sector service being researched.